Operando XAFS insights into the dynamic evolution of Cu-based catalysts for light-promoted CO₂ hydrogenation

Summary

Photothermal catalytic reduction of CO₂ to CO has been regarded as a promising strategy for solar-to-chemical energy conversion, yet the dynamic evolution of active sites under realistic reaction conditions remains poorly understood. Herein, we synthesized CuO/ZnO (CuZnO) catalysts with intimate interfacial contact via co-precipitation followed by calcination method. Under photothermal conditions, CuZnO underwent insitu reconstruction into low-coordinated metallic Cu in close contact with ZnO, delivering a CO generation rate of ∼18.1 mmol g⁻¹ h⁻¹ with a selectivity as high as 99% at 200 °C, outperforming that achieved under purely thermal conditions. Operando photothermal X-ray absorption fine structure (XAFS) spectroscopy revealed the real-time evolution of both Cu and Zn coordination environments in CuZnO, illustrating the formation and stabilization of metallic Cu on ZnO during photothermal catalysis. This study elucidates the mechanisms underlying photothermal synergy in CO₂ reduction and provides mechanistic insights into active-site dynamics under photothermal conditions.

Graphical abstract

Highlights

• •Operando XAFS reveals CuZnO reconstruction into active Cu/ZnO interfaces
• •Photothermal CuZnO gives 18.1 mmol g⁻¹ h⁻¹ CO with >99% selectivity
• •Photothermal operation gives a lower apparent activation energy
• •Photothermal CuZnO shows stable CO generation over 55 h with >99% selectivity

Chemistry; Catalysis; Materials characterization

Introduction

The ever-increasing atmospheric CO₂ concentration driven by human activities has intensified the urgency for sustainable carbon management strategies.¹ Converting CO₂ into value-added chemicals is widely regarded as a promising approach to mitigating carbon emissions and closing the anthropogenic carbon cycle.² Among the various CO₂ utilization routes, hydrogenation of CO₂ to CO via the reverse water-gas shift (RWGS) reaction is particularly attractive because CO is a key feedstock for Fischer-Tropsch synthesis (FTS) and downstream C1 chemistry.^3,4,5,6 However, the high thermodynamic stability and kinetic inertness of CO₂ impose substantial activation barriers, such that appreciable reaction rates are typically achieved only at elevated temperatures. Conventional thermocatalytic RWGS can deliver high CO₂ conversion and selectivity under harsh conditions, yet it often entails significant energy input and, in many scenarios, remains indirectly coupled to fossil-derived heat sources.⁷ In contrast, photocatalysis represents an environmentally benign alternative that can operate under milder conditions, but its practical efficiency is frequently limited by incomplete solar-spectrum utilization, rapid charge-carrier recombination, and sluggish surface kinetics for multi-step CO₂ activation.^8,9,10,11

Photothermal catalysis has recently emerged as a compelling strategy that integrates photon harvesting with thermal activation, enabling CO₂ conversion under comparatively moderate macroscopic conditions.^12,13,14,15 In such systems, plasmonic or broadband-absorbing components convert incident light into localized heat and, in some cases, energetic charge carriers, thereby accelerating surface kinetics and effectively lowering apparent reaction barriers.^16,17 Among plasmonic materials, Cu is particularly appealing as an earth-abundant alternative to noble metals because it can exhibit localized surface plasmon resonance (LSPR) in the visible-near-infrared region when appropriately nanostructured, enabling efficient broadband solar absorption and photothermal conversion.^18,19,20,21 Beyond its optical advantages, Cu is also catalytically relevant for hydrogenation-related chemistry.^22,23,24 Notably, under CO₂ hydrogenation conditions, Cu-based catalysts might undergo in situ reduction and structural reconstruction, generating under-coordinated metallic Cu sites and interfacial motifs that may differ markedly from the as-prepared precatalyst.^25,26 In this context, Cu/ZnO represents one of the most important and widely studied catalyst systems for CO₂ hydrogenation. Its catalytic performance has generally attributed to the synergistic interaction between Cu and ZnO.²⁷ Recent studies have further shown that light irradiation can markedly enhance the activity of Cu/ZnO, with a reported 10.4-fold increase compared with the non-irradiated thermocatalytic process, owing to the simultaneous excitation of Cu plasmons and the ZnO band gap.²⁸ In addition, the synergistic mechanism and structural evolution of Cu/ZnO during photo-assisted methanol synthesis have been investigated in previous studies.²⁹ Nevertheless, current understanding of the active sites in Cu/ZnO-based photothermal catalysis remains limited, because most conclusions rely on ex situ characterization or theoretical predictions. As a result, the dynamic evolution of the catalyst under realistic working conditions is still difficult to capture directly. In particular, understanding how Cu evolves under working conditions, particularly under illumination where reduction and reconstruction may be accelerated, is essential for identifying the true active structure and rationalizing photothermal enhancement.

In this work, we employed operando photothermal X-ray absorption fine structure (XAFS) spectroscopy to directly visualize the atomic-scale evolution of a Cu-based CuZnO catalyst during CO₂ hydrogenation. We revealed a cooperative reconstruction process in which Cu species were reduced to metallic Cu nanoparticles, while the ZnO support largely retained its framework but underwent local coordination modulation under combined light and heat, together forming an active Cu-ZnO interfacial architecture. The in situ reconstructed catalysts delivered a CO formation rate of ∼18.1 mmol g⁻¹ h⁻¹ with high selectivity (>99%) at 200 °C under light irradiation, outperforming the activity achieved under purely thermal conditions. These operando insights establish a clear structure-activity relationship and provide mechanistic guidance for designing efficient non-noble-metal photothermal catalysts for CO₂ conversion.

Results

Material synthesis and characterization

The CuZnO catalyst with a Cu:Zn ratio of 1:2 was prepared by co-precipitation followed by calcination, yielding a composite with intimate Cu-Zn oxide contact. The resultant CuZnO was first investigated by X-ray diffraction (XRD). As shown in Figure S1, the diffraction peaks at 35.5°, 38.6°, and 48.8° were assigned to the characteristic reflections of CuO, corresponding to the (−111), (111), and (−202) facets (PDF#80-0076).³⁰ In addition, the peaks at 31.7°, 34.3°, 36.2°, and 56.5° were attributed to ZnO, relating to the (100), (002), (101), and (110) reflections (PDF#80-0074).³¹ This confirms the coexistence of CuO and ZnO. To obtain more detailed structural and morphological information, scanning electron microscopy (SEM) and transmission electron microscopy (TEM) analyses were conducted. As depicted in Figure S2A, the CuZnO catalyst displayed a flocculent spherical morphology composed of loosely assembled subunits. TEM images further revealed that these spheres were constructed from nanoscale building blocks, forming a porous architecture (Figure S2B). The corresponding energy-dispersive X-ray spectroscopy (EDS) elemental mapping demonstrated a homogeneous distribution of Cu, Zn, and O throughout the composite (Figure S2C). High-resolution TEM (HRTEM) further verified the crystalline nature of the composite, clearly resolving lattice fringes with interplanar spacings of 2.47 Å and 2.61 Å, corresponding to the (101) and (110) planes of ZnO, respectively. Meanwhile, lattice fringes with a spacing of 2.31 Å are also observed and can be assigned to the (111) plane of CuO, consistent with the XRD results (Figure S2D). Thermogravimetric analysis showed that the CuZnO sample was thermally stable below ∼350°C (Figure S3).

Catalytic activity

The CuZnO catalyst was subsequently evaluated for CO₂ hydrogenation in a continuous-flow microreactor (Figure S4). As shown in Figure 1A, under purely thermal operation in the dark, the CuZnO catalyst exhibited negligible activity at 120°C, with no detectable reduction products. Upon increasing the temperature to 160°C, CO formation became observable, although the rate remained very low (2.49 μmol g⁻¹ h⁻¹). As the temperature was further increased from 160°C to 280°C, the CO formation rate gradually increased but was still modest, reaching 39.60 μmol g⁻¹ h⁻¹ at 200°C and 0.66 and 4.54 mmol g⁻¹ h⁻¹ at 240°C and 280°C, respectively. These results indicate that thermal energy alone is insufficient to effectively overcome the kinetic barriers associated with CO₂ hydrogenation on CuZnO under the present conditions. To overcome this limitation, light irradiation was introduced to probe the photothermal catalytic behavior of CuZnO and to evaluate whether light-heat synergy could further enhance CO₂ hydrogenation. Remarkably, upon light irradiation at an intensity of 1.15 W cm⁻², the catalytic activity of CuZnO was dramatically enhanced across the entire temperature range investigated. Notably, at 200°C, the CO generation rate increased sharply from 39.60 μmol g⁻¹ h⁻¹ under purely thermal conditions to 18.1 mmol g⁻¹ h⁻¹ under photothermal conditions, corresponding to an enhancement of more than two orders of magnitude. This performance is comparable to that of representative advanced catalysts reported for photothermal CO₂ hydrogenation (Table S1). Similar promotion was observed at higher temperatures, underscoring the decisive role of light irradiation in boosting catalytic performance. To further quantify the contribution of light, the CO production rate was measured as a function of light intensity at a fixed temperature of 200°C. As depicted in Figure 1B, the CO formation rate increases monotonically (approximately linearly) with increasing light intensity from 0.90 to 1.15 W cm⁻², evidencing a strong light-dependent response and suggesting that the activity enhancement cannot be attributed solely to conventional bulk heating. Kinetic analysis based on Arrhenius plots (Figure 1C) revealed a substantially reduced apparent activation energy under photothermal conditions (25.9 kJ mol⁻¹) compared with thermal operation (125.1 kJ mol⁻¹), indicating that light-promoted reconstruction and enhanced activity under photothermal conditions. The long-term stability of the CuZnO catalyst was evaluated at 200°C (Figure 1D). Under photothermal conditions, the catalyst sustained a CO formation rate of ∼15–18 mmol g⁻¹ h⁻¹ with high selectivity (>99%) over 55 h of continuous operation, whereas negligible activity was observed under thermal conditions at the same set temperature. Collectively, these results demonstrate the excellent durability of CuZnO under irradiation and confirm that light-heat synergy enables efficient and sustained CO₂ hydrogenation.

Figure 1.

Catalytic activity investigation

(A) CO production rates of the CuZnO catalyst as a function of reaction temperature under purely thermal and photothermal conditions (at a light density of 1.15 W cm⁻²).

(B) Dependence of CO production rate on light intensity at a fixed reaction temperature of 200°C under photothermal conditions.

(C) Arrhenius plots derived from reaction rates under thermal and photothermal conditions, showing the apparent activation energies for CO₂ hydrogenation.

(D) Long-term stability test of the CuZnO catalyst for CO₂ hydrogenation at 200°C under thermal and photothermal conditions (at a light density of 1.15 W cm⁻²).

Operando photothermal XAFS investigation

To probe the active components and the structural evolution of CuZnO during the reaction, operando photothermal XAFS measurements were conducted (Figure S5).³² As displayed in Figure 2A, the Cu K-edge XANES of the pristine CuZnO closely resembles that of CuO, indicating that Cu was predominantly present as Cu²⁺ prior to reaction. Upon heating to 200°C under a CO₂/H₂ feed in the dark (0 min), the absorption edge shifted slightly to lower energy and the white-line intensity decreased, suggesting that Cu reduction was already initiated at 200°C. With increasing time on stream (15–60 min), this trend became progressively more pronounced, consistent with a continuous conversion of Cu-O environments toward more reduced Cu species. Notably, once light irradiation was introduced after 60 min, the Cu K-edge spectrum rapidly approached that of Cu foil, evidencing an accelerated and deeper Cu²⁺ → Cu⁰ transformation under photothermal conditions. The corresponding Fourier transformed EXAFS spectra further corroborated the previous electronic-state evolution. As shown in Figure 2B, the pristine CuZnO sample exhibited a prominent contribution in the low-R region (ca. 1–2 Å, uncorrected for phase shift), which was characteristic of Cu-O coordination in oxidized Cu species. Notably, already at 200°C in the dark (0 min), the Cu-O feature began to attenuate, indicating that the local Cu-O environment was being disrupted at the very early stage of reaction. With increasing time on stream in the dark (15–60 min), the Cu-O contribution gradually diminished, accompanied by the emergence and progressive growth of a higher-R feature (ca. 2–3 Å), attributable to Cu-Cu coordination, suggesting the onset of incipient metallization and the formation of metallic Cu domains. Upon switching on light irradiation after 60 min, the Cu-Cu peak intensity increases markedly and then stabilizes (light-0/15/30 min), while the Cu-O contribution is further suppressed, confirming that illumination not only accelerated Cu reduction but also promoted the rapid establishment and stabilization of the metallic Cu-Cu local structure. Quantitative EXAFS fitting was carried out to extract the coordination parameters, and the detailed fitting procedures and coordination numbers are summarized in the supplemental information (Figure S6 and Table S2).

Figure 2.

Operando XAFS analysis

(A) Time-resolved XANES collected in the dark (0–60 min) and after light irradiation (0–30 min), together with CuO and Cu foil references.

(B) Corresponding FT-EXAFS spectra (k³-weighted magnitude) showing the evolution from Cu-O to Cu-Cu coordination.

(C) WT-EXAFS maps showing the gradual decrease of Cu-O contributions and the concurrent increase of Cu-Cu scattering under photothermal conditions.

To further distinguish the scattering contributions and avoid ambiguity arising from peak overlap in FT space, we performed wavelet-transform EXAFS (WT-EXAFS) analysis (Figure 2C). The CuO reference displayed its main WT intensity in the lower-k region, consistent with dominant Cu-O scattering, whereas Cu foil showed a strong feature in the higher-k region, characteristic of Cu-Cu scattering. For the operando CuZnO sample, the WT maps revealed a clear and continuous transition: during the dark period at 200°C (0–60 min), the low-k Cu-O related intensity progressively weakened while a high-k Cu-Cu related feature became increasingly evident, directly visualizing the gradual replacement of Cu-O coordination by Cu-Cu coordination. After light irradiation was introduced, the high-k Cu-Cu feature was further intensified and became more like that of Cu foil, whereas the low-k Cu-O signature was largely suppressed, demonstrating that metallic Cu-Cu coordination becomes the dominant structural motif under photothermal conditions. Taken together, the FT-EXAFS and WT-EXAFS results are fully consistent with the XANES evolution in Figure 2A, confirming that Cu reconstruction is initiated even at 200°C in the dark and is strongly accelerated and deepened upon light illumination. To determine whether light irradiation alone can induce catalyst evolution, we performed operando XAFS measurements under light-only conditions without external heating and compared the results with those obtained under photothermal conditions. As shown in Figure S7A, the Cu K-edge XANES spectra remain highly similar to those of CuO during 0–30 min of light irradiation, with no obvious change relative to the pristine CuZnO sample. Likewise, the FT-EXAFS spectra (Figure S7B) show no attenuation of the Cu-O coordination peak in the low-R region and no emergence of a Cu-Cu coordination peak in the high-R region. These results demonstrate that light alone cannot induce Cu²⁺ reduction or Cu-species reconstruction. Operando Zn K-edge XAFS indicated that the ZnO component largely maintained its chemical state under working conditions. The Zn K-edge XANES line shape remained characteristic of Zn²⁺ throughout the experiment, with no signatures of metallic Zn or alternative Zn-containing phases, suggesting that the ZnO framework was largely retained. Notably, the corresponding FT-EXAFS exhibited a clear attenuation of the Zn-O contribution, and quantitative EXAFS fitting revealed a decrease in the average Zn-O coordination number as the reaction proceeds (Figures S8 and S9, and Table S3). These results indicated the generation of low-coordinated Zn sites in ZnO under photothermal RWGS conditions, which would modulate the interfacial adsorption/activation of CO₂-derived species.

To further validate the operando XAFS observations, ex situ characterizations were performed on the catalyst after the photothermal CO₂ hydrogenation reaction. TEM was employed to examine the morphology and local structure of the reconstructed catalyst. As shown in Figure 3A, the post-reaction CuZnO catalyst retained its overall flocculent spherical morphology, indicating that the photothermal reconstruction process did not induce noticeable sintering or structural collapse. High-resolution TEM (Figure 3B) revealed well-defined lattice fringes with an interplanar spacing of approximately 0.18 nm, which can be assigned to the (200) plane of metallic Cu. These Cu nanodomains are uniformly distributed on the ZnO matrix with an average particle size of ∼7 nm. The corresponding EDS elemental mapping (Figure 3C) further confirmed the homogeneous spatial distribution of Cu, Zn, and O throughout the spent catalyst, excluding large-scale Cu aggregation or phase separation during photothermal operation. The phase composition of the reconstructed catalyst was also examined by XRD. As displayed in Figure 3D, the XRD pattern of the post-reaction sample is dominated by diffraction peaks associated with ZnO, while no distinct reflections corresponding to crystalline CuO are observed. Meanwhile, metallic Cu-related peaks are either absent or significantly broadened, indicating that the reduced Cu species are highly dispersed and exist as nanosized domains rather than bulk metallic Cu. This result is fully consistent with the low Cu-Cu coordination number derived from operando EXAFS analysis. In addition, UV-vis diffuse reflectance spectroscopy reveals a marked enhancement of light absorption in the visible-near-infrared region after the photothermal reaction. This broad absorption feature can be attributed to the LSPR of metallic Cu nanoparticles (Figure S10), providing optical evidence for the in situ formation of plasmonic Cu species and their contribution to photothermal catalysis.^28,33 To elucidate the role of ZnO in the CuZnO catalyst, we investigated CO₂ hydrogenation at 200°C under different irradiation wavelength ranges. As shown in Figure S11, the product formation rate is 2.2 mmol g⁻¹ h⁻¹ under 300–420 nm irradiation, suggesting that ZnO contributes to the reaction through its intrinsic photocatalytic response in the UV region. Under 420–2500 nm irradiation, the production rate increases to 12.6 mmol g⁻¹ h⁻¹, accounting for 77.6% of the CO yield under full-spectrum irradiation. This result indicates that, although ZnO is active under UV light, the major catalytic activity under broad-spectrum irradiation originates from the photothermal effect of the reconstructed CuZnO catalyst, in which metallic Cu plays the dominant role in solar-light harvesting and energy conversion. To probe the electronic consequence of this reconstruction, photocurrent (PC) and electrochemical impedance spectroscopy (EIS) measurements were further performed on the fresh and spent CuZnO catalysts (Figure S12). The spent catalyst exhibits photocurrent intensity 4.8 times higher than that of the fresh sample, indicating more efficient separation and transfer of photogenerated charge carriers after reconstruction. Consistently, the EIS spectra show a smaller semicircle diameter for the spent catalyst, suggesting reduced charge-transfer resistance and enhanced interfacial charge migration. These results further support that the photothermal reaction induces favorable structural evolution in CuZnO, which improves charge transfer and contributes to enhanced catalytic performance.

Figure 3.

Crystalline structure of the CuZnO catalyst after the photothermal reaction

(A) TEM image.

(B) HRTEM image.

(C) SEM-EDS elemental mapping images showing the spatial distribution of Cu, Zn, and O in the reconstructed CuZnO catalyst after the photothermal reaction.

(D) XRD patterns of CuZnO before and after the photothermal reaction.

In situ FTIR analysis

To gain mechanistic insight into photothermal CO₂ hydrogenation on the CuZnO catalyst, in situ FTIR spectroscopy was employed to monitor surface intermediates during a temperature ramp from room temperature to 200°C, followed by light irradiation at 200°C. As shown in Figure 4A, upon heating under flowing CO₂/H₂, distinct vibrational features gradually develop in the fingerprint region, indicating the formation of adsorbed CO₂-derived species. The strong band around ∼2340 cm⁻¹ is attributed to CO₂-related vibrations, while broad features in the 3200–3600 cm⁻¹ region are associated with surface –OH/H₂O species.³³ After reaching 200°C, switching on light irradiation induces noticeable changes in the intensities of surface bands, suggesting that illumination modulates the adsorption reaction dynamics of key intermediates under working conditions. A magnified view of the 1100–1800 cm⁻¹ region (Figure 4B) shows bands attributable to carbonate/bicarbonate species (b-CO₃∗), together with hydrogenated C1 intermediates such as formate (HCOO∗) and carboxyl-type species (COOH∗) (as indicated in the figure).^33,34,35 During the temperature ramp, carbonate/bicarbonate-related bands become increasingly pronounced, consistent with enhanced CO₂ adsorption/activation on the oxide-containing surface. With further evolution under reaction conditions and upon illumination at 200°C, the relative intensities of these carbonate/formate/carboxyl bands change dynamically, indicating accelerated formation and consumption of CO₂-derived intermediates under photothermal operation. Together with the Zn K-edge EXAFS evidence for a decreased Zn-O coordination number (i.e., the generation of under-coordinated Zn sites) and the Cu K-edge results indicating the formation of Cu⁰, these FTIR observations are consistent with an interfacial microenvironment that stabilizes reactive CO₂-derived species and facilitates their hydrogenation toward CO in the RWGS pathway. Based on the aforementioned results, a possible reaction mechanism is proposed, involving two possible pathways: one via linearly adsorbed CO₂ species and the other via bidentate formate-like intermediates.

Figure 4.

In situ FTIR spectra of CuZnO during CO₂ hydrogenation

(A) Full-range spectra recorded under flowing CO₂/H₂ during a temperature ramp (r.t. → 200°C) and subsequent irradiation at 200°C (“light on”).

(B) Enlarged spectra in the 1100–1800 cm⁻¹ region. Bands are assigned to surface CO₂-derived species, including CO₂∗, carbonate/bicarbonate (b-CO₃∗), formate (HCOO∗), and carboxyl species (COOH∗).

Reaction pathway of linearly adsorbed CO₂:

H₂ + 2h⁺ → 2H⁺ (Equation 1)

∗CO₂ + e⁻ → ∗CO₂⁻ (Equation 2)

∗CO₂⁻ + H⁺ → ∗COOH (Equation 3)

∗COOH + e⁻ + H⁺ → ∗CO + H₂O (Equation 4)

∗CO → CO (Equation 5)

Reaction pathway of bidentate formate species:

H₂ → 2H∗ (Equation 6)

∗CO₂ + 2V_o + H → b-COOH∗ (Equation 7)

b-COOH∗ → ∗CO + ∗OH (Equation 8)

∗OH + H → H₂O (Equation 9)

∗CO → CO (Equation 10)

Discussion

In conclusion, this study elucidates the dynamic evolution of CuZnO catalysts under photothermal RWGS conditions, with a particular emphasis on insights obtained from operando photothermal XAFS. Operando Cu and Zn K-edge XAFS directly capture the in situ reduction of Cu²⁺ to low-coordinated metallic Cu and the simultaneous coordination decrease of ZnO during reaction, revealing a cooperative reconstruction process driven by light-heat synergy. Complementary ex situ characterizations confirm that the reconstructed catalyst consists of highly dispersed plasmonic Cu nanoparticles strongly interacting with defect-rich ZnO. In situ FTIR further suggests that these interfacial sites promote CO₂ activation through carbonate- and carboxyl-mediated pathways. As a result, the in situ formed Cu/ZnO catalyst exhibits high CO productivity, near-unity selectivity, and excellent stability with substantially reduced activation energy. This work highlights operando XAFS as a powerful tool for uncovering active-site dynamics in photothermal catalysis and guiding the rational design of efficient non-noble-metal catalysts for CO₂ conversion.

Limitations of the study

Although this study provides operando insight into the dynamic reconstruction of CuZnO catalysts under photothermal CO₂ hydrogenation conditions, several limitations should be noted. Although operando XAFS captures the average evolution of Cu and Zn species under working conditions, it cannot fully resolve the spatial heterogeneity of interfacial active sites at the atomic scale. In addition, the proposed reaction mechanism is mainly inferred from operando FTIR together with XAFS results, and the exact elementary steps still require further validation by complementary approaches such as isotopic labeling or theoretical calculations. Moreover, although the results confirm that catalyst reconstruction depends on the synergistic effect of light and heat, the individual contributions of photothermal heating and non-thermal photoinduced effects are not fully separated in the present work.

Resource availability

Lead contact

Further information and other requests should be directed to and will be fulfilled by the lead contact, Guangbo Chen (gchen@mail.ipc.ac.cn).

Materials availability

This work did not generate new unique reagents. All materials and methods used for data generation and analysis are mentioned in the article.

Data and code availability

All data reported in this article will be shared by the lead contact upon request. This article does not report original code. Any additional information required to reanalyze the data reported in this article is available from the lead contact upon request.

Acknowledgments

This work was supported by the National Key R&D Program of China (2022YFA1504601 and 2024YFA1510700), the National Natural Science Foundation of China (22579175), and the Natural Science Foundation of Zhejiang Province (LTGS24B060004).

Author contributions

Z.W., data curation, formal analysis, investigation, and methodology; X.H., data curation, formal analysis, and investigation; B.Q., investigation and methodology; X.Z., investigation and methodology; J.M., formal analysis and investigation; R.F., investigation; M.S., supervision and writing – review and editing; G.C., conceptualization, supervision, and writing – review and editing; F.S., conceptualization, methodology, supervision, writing – original draft, and writing – review and editing.

Declaration of interests

The authors declare no conflict of interest.

STAR★Methods

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Chemicals, peptides, and recombinant proteins
Cu(NO3)2·3H2O (99%)	Aladdin	CAS: 10031-43-3
Zn(NO3)2·6H2O (99%)	Aladdin	CAS: 10196-18-6
Na2CO3 (99%)	Aladdin	CAS: 497-19-8
Software and algorithms
Origin	OriginLab	https://www.originlab.com/
Other
X-ray diffraction	Rigaku	Rigaku MiniFlex 600
Scanning electron microscopy	Zeiss	Zeiss Sigma 500 VP
Transmission electron microscopy	JEOL	JEOL JEM-2100
X-ray photoelectron spectroscopy	Thermo Fisher Scientific	Thermo Scientific ESCALAB 250
UV-vis diffuse reflectance spectroscopy	Shimadzu	Shimadzu UV-2600i

Experimental model and study participant details

There are no experimental models (animals, human participants, plants, microbe strains, cell lines, primary cell cultures) used in the study.

Method details

Synthesis of CuZnO

CuZnO catalysts with a Cu/Zn molar ratio of 1:2 were prepared via a co-precipitation method. An aqueous solution containing 0.05 M Cu(NO₃)₂·3H₂O and 0.10 M Zn(NO₃)₂·6H₂O was used as the metal precursor, and a 0.15 M Na₂CO₃ aqueous solution served as the precipitant. Both solutions were simultaneously added dropwise into 200 mL of deionized water at 70°C under vigorous stirring. After the addition of 100 mL of the metal precursor solution, the resulting suspension was aged at 70°C for 2 h, cooled to room temperature, and filtered. The collected precipitate was dried overnight at 110°C and calcined in air at 350°C for 4 h.

Characterizations

Powder X-ray diffraction (XRD) patterns were collected on a Rigaku MiniFlex 600 diffractometer using Cu Kα radiation. Transmission electron microscopy (TEM) and high-resolution TEM (HRTEM) were performed on a JEOL JEM-2100 microscope. Scanning electron microscopy (SEM) images were obtained using a Zeiss Sigma 500 VP field-emission SEM. X-ray photoelectron spectroscopy (XPS) was conducted on a Thermo Scientific ESCALAB 250 spectrometer with Al Kα radiation. UV-vis diffuse reflectance spectra (UV-vis DRS) were recorded on a Shimadzu UV-2600i spectrophotometer.

Catalytic test

The photothermal CO₂ hydrogenation performance of the CuZnO catalysts was evaluated in a continuous-flow microreactor (Beijing Zhongjiao Jinyuan Technology Co., Ltd.). For each test, 50 mg of catalyst was loaded into the flat sample bed of a quartz tube reactor (open at both ends) with quartz wool packed at the bottom. The reactor was placed in the heating zone of the furnace, and a CO₂/H₂ mixture (molar ratio 1:3; flow rates of 2 and 6 mL min^-1, respectively) was fed continuously. Thermal catalysis was first examined at 200°C for 1.5 h under dark conditions, followed by photothermal catalysis under Xe lamp irradiation at the same temperature to evaluate the synergistic effect of light and heat. For temperature-dependent activity measurements, the reaction was conducted at 120, 160, 200, 240, and 280°C under dark and irradiated conditions, respectively. The gaseous products were analyzed using a GC-7920 gas chromatograph (GC) equipped with a flame ionization detector (FID). The GC contained two analytical channels. Channel FID1, fitted with a ZKAT-PLOT Pora Q column, was used for hydrocarbon/oxygenate analysis (e.g., CH₄ and CH₃OH, with no oxygenates and higher hydrocarbon detected in our system). Channel FID2, fitted with a TDX-01 column and a methanizer, was used for CO and CO₂ quantification. Nitrogen (N₂) was used as the carrier gas for both channels.

$$\text{Product}{\mathrm{s}\text{electivity}}_{\left(\text{CO}\right)}=\frac{R_{\text{CO}}}{R_{\text{CO}}+R_{\mathrm{C}{\mathrm{H}}_4}}\times 100\%$$

Operando XAFS analysis

XAFS measurements were conducted at beamline BL20U1 of the Shanghai Synchrotron Radiation Facility (SSRF, 3.5 GeV, 200 mA). Schematic diagram of the operando photothermal XAFS experimental setup was reported in previous work.³² Briefly, the temperature control system uses a Yudian AI-828 temperature controller, with a heating mode consistent with that of the muffle furnace and tubular furnace used in the off-line experiments. The heating process is regulated by a programmed temperature profile, with a ramping rate below 10°C min^-1. Both isothermal and stepwise temperature programs can be applied depending on the experimental design. In addition, an external PT100 thermal resistor is used for temperature compensation, giving an internal compensation accuracy of ±1°C. The gas flow system is regulated by mass flow controllers, which allow precise control of the flow rate in each channel as well as flexible adjustment of gas composition and ratios. Cu and Zn K-edge spectra were acquired using a fixed-exit Si (111) double-crystal monochromator and calibrated with Cu and Zn metal foils. Operando spectra were collected in transmission mode using a custom in situ cell with Kapton windows, gas ports, and a thermocouple positioned near the catalyst bed. The cell was fed with the same CO₂/H₂ mixture as used in catalysis tests. Photothermal conditions were realized by irradiating the catalyst bed through an optical window while maintaining the bed temperature at 200°C. XANES and EXAFS data were processed with standard procedures (background subtraction, normalization, k³ weighting) and fitted in R space using FEFF-based models. Details of fitting paths and constraints are provided in the Supporting Information.

Insitu FTIR analysis

Operando FTIR spectra were collected in a dedicated diffuse-reflectance cell under flowing CO₂/H₂. Spectra were recorded during temperature ramps (room temperature to 200°C) and under photothermal irradiation at 200°C. All spectra were referenced to the background collected under inert gas at the corresponding temperature.

Quantification and statistical analysis

No methods were used to determine whether the data met the assumptions of the statistical approach.

Additional resources

Our study has not generated or contributed to a new website/forum or has not been part of a clinical trial.

Published: April 21, 2026